We have initiated a tumor-infiltrating lymphocyte (TIL) therapy program that has now treated over 70 metastatic melanoma patients with an ongoing 45-50% clinical response rate. In parallel, we are performing correlative biomarker studies to identify specific lymphocyte subsets within infused TIL associated with clinical responses. In a recent FACS-based phenotypic screen on over 31 treated patients, we found that a subset of differentiated CD3+CD8+CD45RA-CD62L-CD27-effector-memory (EM) cells expressing B- and T- lymphocyte attenuator (BTLA) stood out to be most strongly associated with clinical response. Other EM markers, such as CD27, CD28, CD25, and PD-1 were not predictive. Although BTLA is classically defined as a negative co-stimulatory molecule on activated T cells, recent data suggests it is also a new marker in the CD8+ CTL differentiation program and that melanoma-specific CD8+ T cells expressing BTLA have a less differentiated, more polyfunctional phenotype in vivo. Preliminary data in our lab has also found that sorted CD8+BTLA+ melanoma TIL have a high proliferative response to IL-2, while their BTLA- counterparts were hyporesponsive and prone to apoptosis. In addition, we have found that once BTLA is lost on T cells, it cannot be re-expressed. These observations have led us to hypothesize that the BTLA+CD8+ TIL subset represents a less differentiated, polyfunctional subset whose persistence in vivo results in enhanced responsiveness to adoptive T cell therapy in melanoma. To address this hypothesis, we propose to further unravel the role of this CD8+BTLA+ melanoma TIL subset in adoptive cell therapy. In Aim #1, we will perform a series of in vitro assays on sorted BTLA+ and BTLA- TIL studying their EM properties (proliferation, survival, response to TCR and cytokine stimulation) and their anti-tumor effector activity. In Aim #2, using high-throughput TCR V? CDR3 region sequencing, we will track the long-term fate of T-cell clones in vivo from the CD8+BTLA+ or CD8+BTLA- sub-populations infused into patients. In Aim #3, we will focus in more on the mechanistic role of the CD8+BTLA+ versus the BTLA- subset in a set of adoptive transfer experiments in NOD-SCID x ?c-/- (NSG) mice, where we will study the relative persistence, tumor infiltration, and anti-tumor effector function of these subsets. These experiments will also test whether blockade of BTLA, using anti-BTLA antibodies, enhances tumor control to address a corollary hypothesis that although BTLA is a marker for more highly functional T cells, it still serves as a negative costimulatory molecule through its interaction with the herpes virus entry mediator (HVEM) expressed on melanoma cells. This project is innovative as it will characterize a novel biomarker in adoptive T-cell therapy whose loss may identify a key turning point in CD8+ CTL differentiation towards a senescent state.